Research Papers
Present new work, evidence, or ideas related to secure and trustworthy machine learning; up to 12 pages of body text.
Security/CORE Unranked
SATML 2026
IEEE Conference on Secure and Trustworthy Machine Learning
The 5th IEEE Conference on Secure and Trustworthy Machine Learning (SaTML 2027) aims to advance the theoretical and practical understanding of vulnerabilities in machine learning, promote robustness in learning systems, and foster a unified scientific community dedicated to trustworthy ML. The conference welcomes research, systematization of knowledge, and position papers that engage substantively with safety, security, privacy, fairness, and verification in ML.
Key deadlines
Verified September 9, 2026Full paper
September 25, 2025
AoE
Conference timeline
Submission and decisions
Full paperKey deadline
September 25, 2025 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Systematization of Knowledge (SoK) Papers
Consolidate and clarify ideas or challenge long-held beliefs in secure and trustworthy ML; must include 'SoK:' in title and be up to 12 pages of body text.
Position Papers
Cover broader issues and visions such as open challenges, societal impact, or educational aspects; must include 'Position:' in title and be 5 to 12 pages of body text.
Research areas in scope
01
Areas of Interest
Novel attacks on machine learningNovel defenses for machine learningSecure and safe machine learning in practiceVerification of algorithms and systemsMachine learning system securityPrivacy in machine learningForensic analysis of machine learningFairness and interpretabilityTrustworthy machine learning in cybersecurity applicationsTrustworthy data curation
Policies worth checking twice
- Submissions must be double-blind: authors' names and institutions must be omitted, and self-citations must be in third person.
- Authors must provide ORCIDs and confirm Author Certification by the abstract registration deadline.
- Papers must not substantially overlap with papers published, accepted, or under review elsewhere; double submissions are grounds for reporting to IEEE.
- All submissions must include an 'Open Science' section describing artifacts or reasons for non-sharing; artifacts must be anonymized and uploaded to a repository within 3 days of submission.
- Accepted papers must share final artifacts on Zenodo; acceptance is conditional on artifact availability unless a valid reason is provided.
- If LLMs are used, an 'LLM usage considerations' section is mandatory, addressing accountability, transparency, and responsibility.
- Authors must append prior reviews from previous submissions (anonymized) if the paper was previously reviewed elsewhere.
- Submissions must use the official IEEEtran two-column conference template; deviations are grounds for desk rejection.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Official conference siteCall for PapersCall for WorkshopsCall for PapersCall for Workshopsworkshop template file
Last verified September 9, 2026